Tendon proper- and peritenon-derived progenitor cells have unique tenogenic properties
© Mienaltowski et al.; licensee BioMed Central Ltd. 2014
Received: 11 December 2013
Accepted: 30 June 2014
Published: 8 July 2014
Skip to main content
© Mienaltowski et al.; licensee BioMed Central Ltd. 2014
Received: 11 December 2013
Accepted: 30 June 2014
Published: 8 July 2014
Multipotent progenitor populations exist within the tendon proper and peritenon of the Achilles tendon. Progenitor populations derived from the tendon proper and peritenon are enriched with distinct cell types that are distinguished by expression of markers of tendon and vascular or pericyte origins, respectively. The objective of this study was to discern the unique tenogenic properties of tendon proper- and peritenon-derived progenitors within an in vitro model. We hypothesized that progenitors from each region contribute differently to tendon formation; thus, when incorporated into a regenerative model, progenitors from each region will respond uniquely. Moreover, we hypothesized that cell populations like progenitors were capable of stimulating tenogenic differentiation, so we generated conditioned media from these cell types to analyze their stimulatory potentials.
Isolated progenitors were seeded within fibrinogen/thrombin gel-based constructs with or without supplementation with recombinant growth/differentiation factor-5 (GDF5). Early and late in culture, gene expression of differentiation markers and matrix assembly genes was analyzed. Tendon construct ultrastructure was also compared after 45 days. Moreover, conditioned media from tendon proper-derived progenitors, peritenon-derived progenitors, or tenocytes was applied to each of the three cell types to determine paracrine stimulatory effects of the factors secreted from each of the respective cell types.
The cell orientation, extracellular domain and fibril organization of constructs were comparable to embryonic tendon. The tendon proper-derived progenitors produced a more tendon-like construct than the peritenon-derived progenitors. Seeded tendon proper-derived progenitors expressed greater levels of tenogenic markers and matrix assembly genes, relative to peritenon-derived progenitors. However, GDF5 supplementation improved expression of matrix assembly genes in peritenon progenitors and structurally led to increased mean fibril diameters. It also was found that peritenon-derived progenitors secrete factor(s) stimulatory to tenocytes and tendon proper progenitors.
Data demonstrate that, relative to peritenon-derived progenitors, tendon proper progenitors have greater potential for forming functional tendon-like tissue. Furthermore, factors secreted by peritenon-derived progenitors suggest a trophic role for this cell type as well. Thus, these findings highlight the synergistic potential of including these progenitor populations in restorative tendon engineering strategies.
Native repair of tendon following an injury such as a rupture or a tear is achieved through intrinsic and extrinsic mechanisms. When an injury occurs, leukocytes and fibroblasts migrate into the lesion early in repair . Recently, it was determined that many of these fibroblasts originate from the paratenon and are thus an extrinsic source of cells involved in repair . Moreover, cells from within the endotenon are thought to serve as an intrinsic source for repair, including tenocytes. Besides terminally differentiated cells such as leukocytes extrinsically and tenocytes intrinsically, multipotent progenitor populations also have been demonstrated to exist within the tendon proper and peritenon (paratenon and epitenon) of the tendon [3, 4]. Progenitor populations derived from the tendon proper and peritenon of the Achilles tendon are enriched with distinct cell types that are distinguished by expression of markers of tendon and vascular/pericyte origins, respectively . Specifically, isolated progenitors of the tendon proper express greater levels of scleraxis and tenomodulin while progenitors of the peritenon express greater levels of endomucin and prominin 1 . Preliminary studies indicate that progenitor cells can be found within repair tissue of healing patellar tendon injuries . Thus, stem/progenitor cells also may contribute to intrinsic and extrinsic tendon repair mechanisms.
Given the distinctions previously seen in the progenitor cell populations, we hypothesized in this study that tendon proper-derived stem/progenitors and peritenon-derived stem/progenitors possess differing tenogenic properties. Thus, we hypothesized that when incorporated into an in vitro regenerative model, progenitors from each region will respond uniquely and produce constructs with differing tenogenic properties. Tenogenic properties include: collagen-rich protein composition, features of the hierarchical tendon structure, as well as expression of known tenocyte markers [4, 6–12]. To test these hypotheses, isolated peritenon- and tendon proper-derived progenitors were seeded within a contracting fibrinogen/thrombin gel-based tendon construct regenerative tissue model [4, 13–18]. Growth differentiation factor 5 (GDF5) has been shown to influence tenocyte differentiation and potentially tendon development, as well as tendon homeostasis and tendon repair [9, 19–24]. Thus, additionally, we hypothesized that differences in tenogenic properties will also be seen when each seeded progenitor population is treated with GDF5. Finally, because intrinsic and extrinsic mechanisms could work simultaneously within injured tendon via cell-cell crosstalk among cell types, conditioned media experiments were also performed to further define interactions among the two enriched progenitor cell populations and mature tenocytes, or more specifically the outcomes of paracrine signaling.
Ninety-six thirty-day-old (P30) male ScxGFP mice were used for twenty-four total progenitor and tenocyte isolations. Tendons also were dissected from embryonic ScxGFP mice (embryonic day (E)15 to E17) (n = 4) for controls in ultrastructural comparisons and from postnatal day P1 ScxGFP mice (n = 2) for controls in immuno-histochemical analyses of tendon constructs. ScxGFP mice contain the transgenic tendon promoter Scx-driven reporter GFP . The University of South Florida’s Institutional Animal Care and Use Committee specifically approved this study’s protocols (‘USF R3901’).
Mice were euthanized and Achilles tendons with surrounding paratenon tissues were dissected using a sterile technique, and kept on ice in Dulbecco’s phosphate buffered saline (D-PBS, Life Technologies, Benicia, CA, USA) with antibiotics-antimycotics (100 U/ml penicillin, 100 μg/ml streptomycin, 250 ng/ml amphotericin B, Life Technologies). Each biological replicate represents eight Achilles tendons from four mice. Progenitor cells were isolated from the Achilles tendon proper as well as from the associated peritenon (paratenon and epitenon), utilizing a series of enzyme digestions as previously described [3, 4, 26]. Briefly, dissected tendons were first treated with 0.5% type I collagenase (CLS-1, Worthington, Lakewood, New Jersey, USA) and 0.25% trypsin (Life Technologies) in alpha minimum essential medium (alpha-MEM) for 10 minutes at 37°C. Then the surfaces of the tendons were scraped carefully with a rubber policeman to strip away the cells of the peritenon. Peritenon cells were collected in ice-cold alpha-MEM. The remaining tendon tissue was rinsed in Hank’s Balanced Salt Solution, cut into 1-mm3 pieces, and transferred into a solution of 3 mg/ml CLS and 4 mg/ml Dispase II (Roche, Basel, Switzerland). The digestion was undertaken at 90 oscillations per minute in a 37°C water bath for 20 minutes. Then digested tendon proper was transferred to culture media kept on ice, and fresh collagenase/dispase solution was transferred into the tube with the remaining tendon pieces for 10-minute incubations until all the tissue was digested, each time adding the digest to the media (alpha-MEM, 2 mM L-glutamine, antibiotics/antimycotics, 100 μm 2-mercaptoethanol, 20% fetal bovine serum (FBS)). From both digests (peritenon and tendon), cells were strained with a 70 μm cell strainer. Cells were collected by centrifugation for five minutes at 500 g, resuspended in media and counted using a hemocytometer with Trypan blue staining. Tendon and peritenon cells were plated in tissue culture flasks at 40 cells and 320 cells per cm2, respectively, so that adhering cells form segregated colonies within the flasks. Fourteen days into the culture, primary culture isolations of progenitor cell colonies were collected and used for generating tendon constructs or for the conditioned media experiment.
Tenocytes were isolated from the tendon proper by removing the paratenon with enzyme digestion as described above. The remaining tendon tissue was rinsed in Hank’s Balanced Salt Solution, cut into 1-mm3 pieces, and transferred into a type I collagenase/Dispase II solution, also as described above. After digestion, cells were strained with a 70 μm cell strainer, collected by centrifugation for five minutes at 500 g, and resuspended in media. Tenocytes were plated in tissue culture flasks (50,000 cells per T75), and were grown in alpha-MEM with 2 mM L-glutamine, antibiotics-antimycotics, 10% FBS. After fourteen days into the culture, tenocytes also were used in the conditioned media experiment.
After the initial expansion, stem/progenitor cells or tenocytes were trypsinized from flasks and seeded in monolayer culture in 12-well plates (5.0 × 104/well). Cells were maintained in their respective phenotype maintenance media: stem/progenitor (alpha-MEM, 2 mM L-glutamine, antibiotics/antimycotics, 100 μm 2-mercaptoethanol, 20% FBS) [3, 4] or tenocytes (alpha-MEM, 2 mM L-glutamine, antibiotics/antimycotics, 0.15 mM ascorbate-2-phosphate, 10% FBS)  for 24 hours. Then, cells were rinsed with D-PBS and all cells received alpha-MEM, 1% FBS, L-glutamine, and antibiotics. After 18 hours, media were collected and grouped as tenocyte-conditioned (Tn), tendon-proper-stem/progenitor-conditioned (TP), and peritenon-stem/progenitor-conditioned (PERI). Then each cell type was treated with the conditioned media for 48 hours before cultures were terminated. Total RNA was isolated from the cell layer in each well.
After 45 days in culture, tendon constructs were processed for transmission electron microscopy as described previously for tendons [4, 29, 30]. Biological replicates from each sample group (n = 3 to 8) were sectioned using a Leica Ultracut UCT ultramicrotome, and stained with 2% aqueous uranyl acetate and 1% phosphotungstic acid, pH 3.2. Microscopy was undertaken using a JEOL 1400 Transmission Electron Microscope. Images were digitally captured using a Gatan Orius widefield side mount CC Digital camera. Eight cross-sectional images from non-overlapping regions of the central portion of the construct taken at an instrument magnification of 60,000x were used for measuring diameters of fibrils formed within each biological replicate. Digital images were randomized, masked and transferred to a RM Biometrics-Bioquant Image Analysis System (Memphis, TN, USA) for analysis. Five to eight identical regions of interest (ROIs) per image were used on each image. All fibrils within the predetermined ROIs were measured until at least 100 fibrils were counted per image.
Sequence information for primers used in RT-qPCR
Forward primer sequence
Reverse primer sequence
Tendon construct protein content was analyzed alongside of one day old (P1) ScxGFP mouse Achilles tendons using an immuno-histochemical technique previously described [34, 35]. Briefly, dissected P1 Achilles tendons and 7-day tendon constructs were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS), pH 7.2, for two to four hours at 4°C, then stored in an increasing gradient of sucrose and embedded in Optimum Cutting Temperature (OCT) compound (Tissue-Tek®, Sakura Finetek USA, Torrance, CA, USA). Five-micron cryo-sections were then generated and placed on Superfrost Plus slides (Fisher). Slides were rinsed twice in PBS, quenched with 1% glycine in PBS and blocked with 2% bovine serum albumin (BSA) in PBS at room temperature for one hour. Primary antibodies were applied with 1% BSA in PBS overnight at 4°C and then rinsed twice with PBS containing 0.2% Tween-20. Secondary antibodies were applied with 1% BSA in PBS at room temperature for one hour and then slides were rinsed with PBS containing 0.2% Tween-20 and PBS. Coverslips were applied to slides after application of Vectashield® mounting solution with 4′,6-diamidino-2-phenylindole (DAPI) (1:1 H-1000 to H-1200, Vector Laboratories, Inc., Burlingame, CA, USA) used as a nuclear marker. Negative control samples were incubated identically, except without primary antibody. The primary antibodies that were used included: 10 μg/ml rabbit anti-mouse collagen I (AB765P, EMD Millipore, Corp., Billerica, MA, USA), 1:200 rabbit anti-mouse collagen III (LB-1393, Cosmo Bio USA, Carlsbad, CA, USA), and 1:100 rabbit anti-mouse collagen V (Myriad Genetic Laboratories, Inc., Salt Lake City, UT 84108, USA) . The secondary antibody was an Alexa Fluor 568-conjugated goat anti-rabbit immunoglobulin G (IgG) (Molecular Probes, Eugene, OR, USA) used at 1:200. Images were captured using a DM5500 upright microscope system (Leica) with conditions and integration times set to facilitate comparisons between samples.
These studies utilized both general distribution descriptions as well as tests of statistical significance. Features of the fibril diameter distributions for each construct group included mean fibril diameters, standard deviations, median diameter, and first and third quartile values. Statistical analyses of fibril diameter distributions were performed using the unpaired two-tailed t-test comparing mean fibril diameters of each image analyzed for each group (E15-17, TP, TP + GDF5, PERI, PERI + GDF5). RT-qPCR results for progenitor-seeded construct profiles and for expression profiling results from the conditioned media study were analyzed using non-parametric Mann–Whitney-Wilcoxon tests on unpaired samples.
Mean expression fold differences of TP and PERI constructs
TP PS/PERI PS
TP 7/PERI 7
TP 45/PERI 45
Mean expression fold differences of PERI constructs given GDF5 versus TP constructs
TP 7/PERI 7 + GDF5
TP 45/PERI 45 + GDF5
Intrinsic and extrinsic cells, including stem/progenitors, have been described as plausible participants in the repair of tendon tears and ruptures [1, 2]. In this study, we demonstrated that stem/progenitor populations of the tendon proper and peritenon of the Achilles tendon are capable of forming collagen fibril-rich tendon-like structures in vitro that are comparable to embryonic tendons. However, we found that there are several differences in each progenitor population’s ability to regenerate tendon. These in vitro differences provide insight into how these two progenitor populations may behave during native tendon repair or in tendon formation in tissue engineering.
While both progenitor populations produced tendon-like structures when seeded in fibrin gel constructs, as demonstrated in this study and previously , the tendon proper-derived progenitors demonstrated greater levels of tenogenic differentiation by expression of the tendon markers Scx and Tnmd. Even after supplementation with the tenogenic modulator GDF5, expression differences for these tenogenic marker genes still remained unchanged, although comparable levels of matrix assembly gene expression were noted. Expression differences of Scx and Tnmd in these progenitor populations signify the need for better understanding of the stem/progenitor cells involved in tendon repair as well as the markers that may define the origins and, thus, variable differentiation potentials of intrinsic and extrinsic cell populations [37, 38]. Since cells from the peritenon enter into lesions early, these cells are proposed to transdifferentiate into tenocytes . In the tendon regeneration model, peritenon-derived progenitors are not differentiating into tenocytes in a manner comparable to progenitors from within the tendon, as measured by Scx and Tnmd expression and ScxGFP transgenic reporter. Thus, from the in vitro results, it could be stated that not every cell that expresses matrix assembly genes has transdifferentiated into a tenocyte, but instead into another collagen-producing cell.
In the regeneration model, ultrastructural similarities were seen between tendon-like structures developed from peritenon- and tendon proper-derived progenitors. That is, progenitors seeded within the constructs had processes projecting into the matrix that gathered collections of fibrils for further assembly into fibers as in developing tendons. Also, likewise, these fibrils were aligned longitudinally, along the tendon axis. However, diameter distributions varied significantly between structures produced by tendon proper-derived progenitors and peritenon-derived progenitors. Differences in fibril diameters could be the result in differential expression of regulatory molecules, such as SLRPs and FACIT collagens, as well as fibril nucleating Col5a1 and Col11a1 . The addition of GDF5 did stimulate matrix assembly gene expression in peritenon progenitors, subsequently contributing to the corresponding shift in fibril diameter distributions toward larger diameters. If this is the case, then this would suggest that GDF5 might also stimulate matrix assembly gene expression in peritenon progenitors through another tenogenic transcription factor, such as Mohawk, or the early growth response transcription factors Egr1 and Egr2 [14, 27, 39]. Certainly, transcription factors besides Scx are active in stimulating transcription of type I procollagen genes and other matrix assembly genes, even in the tendon [40, 41]. Further studies are required to determine each transcription factor’s involvement in activation of intrinsic and extrinsic cells in repair.
In examining gene expression of tendon markers and matrix assembly early and late in fibrin gel-based construct cultures, gene expression waned to various extents from Day 7 to Day 45, particularly for TP constructs. However, the decrease in expression was not statistically significant, except for Col11a1 and Col14a1, although decreases occur in expression for these two genes in developing tendons [34, 42, 43]. In fact, expression for many matrix assembly genes wanes over time as tendon development ensues, maturation proceeds and homeostasis occurs [6, 9, 34, 42, 44, 45]. Whether or not this decrease in expression represents the construct reaching a homeostatic state is still something to be determined. This suggests that other stimuli might be necessary to further stimulate the constructs as they form tendon-like structures.
In the debate over the roles of intrinsic and extrinsic cell sources in repair, it is believed that extrinsic cells secrete stimulatory factors during repair [1, 46]. Here we demonstrated that peritenon-derived progenitors secrete a factor or factors that bolster expression of tenogenic differentiation markers and matrix assembly genes in tendon proper-derived progenitors and mature tenocytes, thus suggesting another role for peritenon-derived progenitors besides synthesis of provisional matrix during tendon repair. This stimulatory effect also highlights the necessity for considering the interactions of all intrinsic and extrinsic cells involved in tendon repair, besides tendon niche components, breakdown products, as well as inflammatory mediators. In a recent study, the stimulatory effect of tendon cells has also been demonstrated to promote tenogenic differentiation in amniotic epithelial cells in co-culture . Thus, cell-cell signaling could be invaluable for cuing tenogenic differentiation in native repair; likewise, manipulation of this signaling could lead to improvements in regenerative tissue engineering strategies.
This study did have a few limitations. Only one concentration of GDF5 was used in these studies. This concentration was selected because it had been effective for other mesenchymal stromal cell-based studies [23, 27, 48, 49]. Furthermore, while this study focused on GDF5, several other growth differentiation factors have been implicated in tendon development and could also merit further investigation within the model [50–54]. Moreover, further analyses of morphological structure, particularly dynamic tracking techniques, would be useful for understanding cell behavior within the developing constructs as cells undergo tenogenic differentiation and initiate tendon formation [55–57].
It should be noted that prior to the conditioned media study, cells were expanded in media for a two week period and then cells were trypsinized and then re-seeded, which is essentially the first passage. Progenitors were expanded in progenitor selective media [3, 4]. Tenocytes were expanded in media with 10% FBS; supplementation of media with 10% FBS has been demonstrated to be acceptable for tenocyte expansion [11, 58], although phenotypic drift could become a concern after long periods of time or multiple passages . Phenotypic drift related to long-term, multiple passage culture is also a concern for progenitor cells .
Lastly, the progenitors in this study were examined via in vitro models. The tendon-forming capabilities of progenitors were analyzed in the fibrin gel model, and enriched progenitor populations – tendon proper versus peritenon – were examined individually, although in native repair intrinsic and extrinsic sources may interact to influence repair, as was demonstrated by the conditioned media study. The purpose of the conditioned media study was to offer evidence of cell-cell interactions via paracrine effects to highlight the importance of considering more than one cell type within the healing niche. Future studies are planned to discern how this synergistic interaction can be influenced to regenerate tendon-specific structure and function in engineered constructs via co-culture as well as in situ during repair via cell tracking within healing tendon tissue.
We demonstrated that progenitors from the tendon proper and the peritenon are capable of forming collagen-rich structures. While TPs are more suited for regeneration of tendon structure, the addition of GDF5 stimulated matrix assembly for PERIs. Moreover, PERIs secreted tendon-promoting factors that bolster expression of tendon markers in TPs and tenocytes. These findings suggest that progenitors in and around the Achilles tendon possess unique tenogenic differentiation characteristics. Distinctions found in tenogenic differentiation and matrix assembly potentials between these progenitor pools provide insight into possible individual and interactive roles of multiple intrinsic and extrinsic cell populations during tendon repair.
bovine serum albumin
type 1 collagenase
Dulbecco’s phosphate-buffered saline
fibril associated collagens with interrupted triple helices
fetal bovine serum
peritenon-derived progenitor or peritenon progenitor conditioned
region of interest
small leucine-rich proteoglycan
tenocyte-derived or tenocyte-conditioned
tendon proper-derived progenitor or tendon proper progenitor conditioned
alpha minimum essential medium.
This study was funded by NIH grants from the National Institute of Arthritis and Musculoskeletal and Skin Diseases (MJM: AR058027; DEB: AR044745). The authors graciously thank Dr. Ronen Schweitzer for the gift of ScxGFP mice.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.